Deep Learning Based Approach to Quantification of PET Tracer Uptake in Small Tumors

Laura Dal Toso; Elisabeth Pfaehler; Ronald Boellaard; Julia A. Schnabel; Paul K. Marsden

doi:https://doi.org/10.1007/978-3-030-33843-5_17

Deep Learning Based Approach to Quantification of PET Tracer Uptake in Small Tumors

Laura Dal Toso, Elisabeth Pfaehler, Ronald Boellaard, Julia A. Schnabel, Paul K. Marsden

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

1 Citation (Scopus)

Abstract

In Positron Emission Tomography (PET), quantification of tumor radiotracer uptake is mainly performed using standardised uptake value and related methods. However, the accuracy of these metrics is limited by the poor spatial resolution and noise properties of PET images. Therefore, there is a great need for new methods that allow for accurate and reproducible quantification of tumor radiotracer uptake, particularly for small regions. In this work, we propose a deep learning approach to improve quantification of PET tracer uptake in small tumors using a 3D convolutional neural network. The network was trained on simulated images that present 3D shapes with typical tumor tracer uptake distributions (‘ground truth distributions’), and the corresponding set of simulated PET images. The network was tested on unseen simulated PET images and was shown to robustly estimate the original radiotracer uptake, yielding improved images both in terms of shape and activity distribution. The same network was successful when applied to 3D tumors acquired from physical phantom PET scans.

Original language	English
Title of host publication	Machine Learning for Medical Image Reconstruction - 2nd International Workshop, MLMIR 2019, held in Conjunction with MICCAI 2019, Proceedings
Editors	Florian Knoll, Andreas Maier, Daniel Rueckert, Jong Chul Ye
Publisher	Springer
Pages	181-192
Number of pages	12
Volume	11905 LNCS
ISBN (Print)	9783030338428
DOIs	https://doi.org/10.1007/978-3-030-33843-5_17
Publication status	Published - 1 Jan 2019
Event	2nd International Workshop on Machine Learning for Medical Image Reconstruction, MLMIR 2019 held in Conjunction with 22nd International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2019 - Shenzhen, China Duration: 17 Oct 2019 → 17 Oct 2019

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	11905 LNCS

Conference

Conference	2nd International Workshop on Machine Learning for Medical Image Reconstruction, MLMIR 2019 held in Conjunction with 22nd International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2019
Country/Territory	China
City	Shenzhen
Period	17/10/2019 → 17/10/2019

Keywords

Convolutional neural network
PET
Quantification
Reconstruction

Access to Document

https://doi.org/10.1007/978-3-030-33843-5_17

https://link.springer.com/content/pdf/10.1007%2F978-3-030-33843-5_17.pdf

Cite this

Dal Toso, L., Pfaehler, E., Boellaard, R., Schnabel, J. A., & Marsden, P. K. (2019). Deep Learning Based Approach to Quantification of PET Tracer Uptake in Small Tumors. In F. Knoll, A. Maier, D. Rueckert, & J. C. Ye (Eds.), Machine Learning for Medical Image Reconstruction - 2nd International Workshop, MLMIR 2019, held in Conjunction with MICCAI 2019, Proceedings (Vol. 11905 LNCS, pp. 181-192). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11905 LNCS). Springer. https://doi.org/10.1007/978-3-030-33843-5_17

Dal Toso, Laura ; Pfaehler, Elisabeth ; Boellaard, Ronald et al. / Deep Learning Based Approach to Quantification of PET Tracer Uptake in Small Tumors. Machine Learning for Medical Image Reconstruction - 2nd International Workshop, MLMIR 2019, held in Conjunction with MICCAI 2019, Proceedings. editor / Florian Knoll ; Andreas Maier ; Daniel Rueckert ; Jong Chul Ye. Vol. 11905 LNCS Springer, 2019. pp. 181-192 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{7ab758e4e91240e08ad035389e4f26f7,

title = "Deep Learning Based Approach to Quantification of PET Tracer Uptake in Small Tumors",

abstract = "In Positron Emission Tomography (PET), quantification of tumor radiotracer uptake is mainly performed using standardised uptake value and related methods. However, the accuracy of these metrics is limited by the poor spatial resolution and noise properties of PET images. Therefore, there is a great need for new methods that allow for accurate and reproducible quantification of tumor radiotracer uptake, particularly for small regions. In this work, we propose a deep learning approach to improve quantification of PET tracer uptake in small tumors using a 3D convolutional neural network. The network was trained on simulated images that present 3D shapes with typical tumor tracer uptake distributions ({\textquoteleft}ground truth distributions{\textquoteright}), and the corresponding set of simulated PET images. The network was tested on unseen simulated PET images and was shown to robustly estimate the original radiotracer uptake, yielding improved images both in terms of shape and activity distribution. The same network was successful when applied to 3D tumors acquired from physical phantom PET scans.",

keywords = "Convolutional neural network, PET, Quantification, Reconstruction",

author = "{Dal Toso}, Laura and Elisabeth Pfaehler and Ronald Boellaard and Schnabel, {Julia A.} and Marsden, {Paul K.}",

year = "2019",

month = jan,

day = "1",

doi = "https://doi.org/10.1007/978-3-030-33843-5_17",

language = "English",

isbn = "9783030338428",

volume = "11905 LNCS",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer",

pages = "181--192",

editor = "Florian Knoll and Andreas Maier and Daniel Rueckert and Ye, {Jong Chul}",

booktitle = "Machine Learning for Medical Image Reconstruction - 2nd International Workshop, MLMIR 2019, held in Conjunction with MICCAI 2019, Proceedings",

note = "2nd International Workshop on Machine Learning for Medical Image Reconstruction, MLMIR 2019 held in Conjunction with 22nd International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2019 ; Conference date: 17-10-2019 Through 17-10-2019",

}

Dal Toso, L, Pfaehler, E, Boellaard, R, Schnabel, JA & Marsden, PK 2019, Deep Learning Based Approach to Quantification of PET Tracer Uptake in Small Tumors. in F Knoll, A Maier, D Rueckert & JC Ye (eds), Machine Learning for Medical Image Reconstruction - 2nd International Workshop, MLMIR 2019, held in Conjunction with MICCAI 2019, Proceedings. vol. 11905 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11905 LNCS, Springer, pp. 181-192, 2nd International Workshop on Machine Learning for Medical Image Reconstruction, MLMIR 2019 held in Conjunction with 22nd International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2019, Shenzhen, China, 17/10/2019. https://doi.org/10.1007/978-3-030-33843-5_17

Deep Learning Based Approach to Quantification of PET Tracer Uptake in Small Tumors. / Dal Toso, Laura; Pfaehler, Elisabeth; Boellaard, Ronald et al.
Machine Learning for Medical Image Reconstruction - 2nd International Workshop, MLMIR 2019, held in Conjunction with MICCAI 2019, Proceedings. ed. / Florian Knoll; Andreas Maier; Daniel Rueckert; Jong Chul Ye. Vol. 11905 LNCS Springer, 2019. p. 181-192 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11905 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

TY - GEN

T1 - Deep Learning Based Approach to Quantification of PET Tracer Uptake in Small Tumors

AU - Dal Toso, Laura

AU - Pfaehler, Elisabeth

AU - Boellaard, Ronald

AU - Schnabel, Julia A.

AU - Marsden, Paul K.

PY - 2019/1/1

Y1 - 2019/1/1

N2 - In Positron Emission Tomography (PET), quantification of tumor radiotracer uptake is mainly performed using standardised uptake value and related methods. However, the accuracy of these metrics is limited by the poor spatial resolution and noise properties of PET images. Therefore, there is a great need for new methods that allow for accurate and reproducible quantification of tumor radiotracer uptake, particularly for small regions. In this work, we propose a deep learning approach to improve quantification of PET tracer uptake in small tumors using a 3D convolutional neural network. The network was trained on simulated images that present 3D shapes with typical tumor tracer uptake distributions (‘ground truth distributions’), and the corresponding set of simulated PET images. The network was tested on unseen simulated PET images and was shown to robustly estimate the original radiotracer uptake, yielding improved images both in terms of shape and activity distribution. The same network was successful when applied to 3D tumors acquired from physical phantom PET scans.

AB - In Positron Emission Tomography (PET), quantification of tumor radiotracer uptake is mainly performed using standardised uptake value and related methods. However, the accuracy of these metrics is limited by the poor spatial resolution and noise properties of PET images. Therefore, there is a great need for new methods that allow for accurate and reproducible quantification of tumor radiotracer uptake, particularly for small regions. In this work, we propose a deep learning approach to improve quantification of PET tracer uptake in small tumors using a 3D convolutional neural network. The network was trained on simulated images that present 3D shapes with typical tumor tracer uptake distributions (‘ground truth distributions’), and the corresponding set of simulated PET images. The network was tested on unseen simulated PET images and was shown to robustly estimate the original radiotracer uptake, yielding improved images both in terms of shape and activity distribution. The same network was successful when applied to 3D tumors acquired from physical phantom PET scans.

KW - Convolutional neural network

KW - PET

KW - Quantification

KW - Reconstruction

UR - http://www.scopus.com/inward/record.url?scp=85076214752&partnerID=8YFLogxK

UR - https://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85076214752&origin=inward

U2 - https://doi.org/10.1007/978-3-030-33843-5_17

DO - https://doi.org/10.1007/978-3-030-33843-5_17

M3 - Conference contribution

SN - 9783030338428

VL - 11905 LNCS

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 181

EP - 192

BT - Machine Learning for Medical Image Reconstruction - 2nd International Workshop, MLMIR 2019, held in Conjunction with MICCAI 2019, Proceedings

A2 - Knoll, Florian

A2 - Maier, Andreas

A2 - Rueckert, Daniel

A2 - Ye, Jong Chul

PB - Springer

T2 - 2nd International Workshop on Machine Learning for Medical Image Reconstruction, MLMIR 2019 held in Conjunction with 22nd International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2019

Y2 - 17 October 2019 through 17 October 2019

ER -

Dal Toso L, Pfaehler E, Boellaard R, Schnabel JA, Marsden PK. Deep Learning Based Approach to Quantification of PET Tracer Uptake in Small Tumors. In Knoll F, Maier A, Rueckert D, Ye JC, editors, Machine Learning for Medical Image Reconstruction - 2nd International Workshop, MLMIR 2019, held in Conjunction with MICCAI 2019, Proceedings. Vol. 11905 LNCS. Springer. 2019. p. 181-192. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: https://doi.org/10.1007/978-3-030-33843-5_17

Deep Learning Based Approach to Quantification of PET Tracer Uptake in Small Tumors

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Cite this